RELIABILITY

Defintion

The extent to which a measure is consistent

replicability

standardization

Reliability in Experiments

Reliability in Observations

Reliability in Interviews

Reliability in Questionnaires / Tests

can't repeat if the IV is a historical event or if the previous study was too unethical

standardized procedure --> allows another researcher to repeat the experiment to see if the results are consistent

inter-rater/observer reliability - the extent to which 2 observers record the same behavior in the same way

can be improved by agreeing on what fits into each category (having operationalized definitions) before the observation begins

Inter-interviewer reliability - the extent to which 2 interviewers interpret the same participant's responses in the same way

Internal reliability

There is a possibility for subjective questions to lack reliability

because one person's definition of 'agree' on a likert scale may be inconsistent with another person's interpretation of the word 'agree'

External reliability

Split halves technique - looks to find out if scores on the first half of the test are the same as scores on the second half of the test

allows us to know whether the test is internally consistent (measures the same factor from start to finish)

Test-retest reliability - the extent to which scores on a test are similar when Ps are tested on the same measure at multiple points in time